In this study, we have developed a robust cryohistological method that allows imaging of virtually any type of plant cell or tissue while preserving fluorescent protein signals and maintaining excellent cellular and subcellular morphology. This method involves modified fixation of plant tissues (i.e., leaves, stems, and petioles), infiltration in a sucrose gradient, freezing, and collection of cryosections directly onto a cryoadhesive tape. Using this method followed by microscopic analysis, we demonstrated a localized accumulation of green fluorescent protein (GFP) in Nicotiana benthamiana plants agroinfiltrated with the movement-incompetent tobacco mosaic virus–based vector and systemic accumulation of GFP in plants infiltrated with the movement-competent vector. Overall, this simple cryohistological procedure reduced sample preparation time and allowed processing of tissue sections for high-resolution imaging of targeted fluorescent proteins in all plant tissues.

In recent years, the agricultural, medical, and pharmaceutical industries have been increasingly recognizing plants for their potential to produce large amounts of recombinant proteins (1,2). A variety of plant virus–based vector systems for expression of recombinant proteins in plants have been described (3-6). The development of recombinant protein expression platforms was greatly facilitated by the use of fluorescent marker proteins (7), which allowed live cell imaging of plant tissues to rapidly determine whether an expression strategy was feasible or not. However, live cell imaging of intact plants with handheld UV lamps or epifluorescence microscopes suffers from several limitations, namely low sensitivity and resolution, making it difficult or impossible to obtain details of protein expression beneath the epidermal cell layers. With very few exceptions, even with confocal or multiphoton microscopy, live cell imaging typically can only monitor fluorescent protein expression a few cell layers beneath the surface. Some further imaging improvements have been recently reported (8), demonstrating that better optical qualities of living plant leaves can be achieved by infiltration with perfluorodecalin, extending the imaging depth of confocal microscopy to about 50 µm. The optical limitations in plants are primarily due to such factors as absorbance, scattering, and severe spherical aberration resulting from refractive index changes associated with cell walls, cell contents, and air pockets. Ultimately, this makes visualization of expression patterns and protein localization in the vast majority of tissue beneath the plant cell surface problematic, especially with vascular and woody tissues and thick mesophyll layers with abundant chloroplasts.

To alleviate such optical effects, the sectioning or clearing of tissues is typically required. Clearing has been a tried and true method that facilitates deep imaging within plant tissues using both standard histological stains and contemporary probes. This procedure includes extraction of chlorophyll and other cell constituents/cell wall materials, staining of structures of interest, and finally, infiltration of tissue with compounds that ultimately render the sample transparent (9,10). The clearing technique is highly useful for studying plant cell/tissue architecture (11). However, due to the nature of clearing with harsh chemicals, only the most abundant and stable cell constituents are preserved. This severely limits the utility of the method for detecting host's labile molecules and fluorescent proteins and application of various techniques for protein and nucleic acid localization.

For decades, traditional histochemical protocols have proven to be invaluable for imaging internal plant structures. The standard procedure consists of tissue fixation, dehydration, rehydration, and subsequent embedment of the plant material into paraffin or plastic polymers. After sectioning of the embedded tissue, the protein of interest is detected using antibodies labeled with a fluorescent molecule, an enzyme, or gold particles (12-15). Among the drawbacks of conventional histological procedures are the lengthy (up to 1 week) sample preparation and the need to work with relatively small (millimeter size) tissue samples to ensure good infiltration with chemical compounds. In addition, the most commonly used fixatives, such as glutaraldehyde and osmium tetroxide (O_sO4), preserve morphological details but may destroy or interfere with fluorescent signals and the integrity of target protein (16,17). Solvents required for dehydration and resin/wax infiltration can also quench fluorescent protein signals and modify or destroy protein's native folding pattern.

Thus, we modified the conventional protocol for fixation and cryosectioning of plant material to achieve the goals of reliable cryosectioning and excellent morphology. In order to minimize tissue disruption and leaching/loss of fluorescent proteins and other cell constituents due to ice growth and segregation artifacts, we chose to cryoprotect sucrose gradient-infiltrated tissues and used a commercially available cryocompatible tape to facilitate retention and maintain the integrity of the plant tissue. Our straightforward procedure reduced sample preparation time to 3 days, allowed the handling of tissue samples of at least 1 cm² in size, preserved cellular and subcellular morphology details, and permitted detection of fluorescence signals from the green fluorescent protein (GFP) tag expressed in plant cells using the tobacco mosaic virus (TMV)–based expression vector. These observations further suggest that detection of immunofluorescence signals also should be also fundamentally possible, as previously described for cryosections of Medicago truncatula plant root nodules treated with a fluorescently labeled anti-syntaxin (SYP132) antibody (18) and for cryosections of tobacco leaf tissue treated with a fluorescently labeled anti-callose antibody (19), with the caveat that as for any technique/probe/tissue, specific optimization would be required.